Abstract: An extensive field trial was set up in eight forest stands to examine the influence of soil texture (two stands on sand, four on loam to silt loam, two on clay), machine mass (light, heavy) and traffic intensity (one and five skidding cycles) (i.e. pass back and forth on the skid trail) on soil compaction after mechanized harvesting. Dry bulk density (BD), penetration resistance (PR), micro-topography and soil carbon dioxide (CO2) concentration were applied as response variables for soil compaction. Significant effects on BD were nearly absent (<7% increase) and occurred occasionally for PR (60โ70% increase, up to 150% on clay soils). Especially for loam to silt loam and clay soils, this was in contrast with the expectation. The negligible compaction degrees for loam to silt loam are attributed to high initial compaction levels that prevented further compaction, as was found by general linear modelling (GLM) for both BD and PR. For clay soils the small compaction degrees can be explained by the high water contents that result in plastic deformation instead of strong compaction degrees, as was confirmed by the micro-topographical measurements. GLM also revealed a significant impact of machine mass (BD) and soil water content (BD, PR) on the compaction degree. Soil texture, traffic intensity and position in relation to the wheel tracks generally turned out to have an insignificant influence. With regard to clear interactions the influence of traffic intensity depends on the position in relation to the wheel tracks and the machine that was used (PR). In contrast with BD and PR, soil CO2 concentration, measured in a forest stand on sand, showed significant increases within and between wheel tracks, even after one skidding cycle. Although soil compaction degrees were small to negligible, machine passes apparently had a strong negative impact on pore continuity. CO2 concentration seems to be a more sensitive and thus better indicator for soil compaction. [Copyright &y& Elsevier]